sales-infludata

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill's documentation (references/platform-guide.md and SKILL.md examples) explicitly describes searching creator profiles, social listening, UGC/comment ingestion, and "semantic comment analysis"/"shitstorm detection" (AI reads and categorizes comments), which means the agent is expected to consume untrusted, user-generated social media content that can influence decisions like campaign actions.